Slashdot Mirror
The Problems With Porting Games
mr_sifter writes "There's a large lexicon of monosyllabic, four-letter words for describing something you don't like — but only PC gamers use the word 'port' with such a fervent degree of repulsion. Common complaints about console ports include meager graphics options, dodgy third-person camera angles, poorly-thought-out controls and sparsely distributed save points. In this feature, Bit-tech talks to developers of games such as Dead Space, Red Faction and Tales of Monkey Island to find out why porting games between the three major consoles and the PC is so difficult. Radically different CPU, graphics and memory architectures play their part, as do the differences in control methods and the rules Microsoft, Sony and Nintendo set about how games should work on their systems."
It's very important to point out that the porting task has everything to do with where you start. The PC is simply not the best development environment anymore, the Xbox 360 is-- and even Carmack would agree with me, here. You can get a game going really fast on 360, then it's a bit less difficult to go to PC. We can call this the best case scenario. Rapid time to market with superior development tools on 360 with familiar API's for cross-platform development on PC, along with similar TCR requirements between GFW and 360.
Let's say you started on the PS3, though. Maybe you took the time to learn the architecture and really take advantage of the cell architecture, so your game is basically hardcoded around the flexible pipeline and mass pararllelization, now it does things that even PC games cannot. Porting it to the 360 might not be so bad, but going to the PC is going to be a rough letdown. It feels like a dog when porting a console game.
So maybe your game started out nicely organized and clean in design, but in that last few months before release while your publisher is driving you up a wall to release, you're going to have so many hacks and messy revisions to the model to ship within your ridiculous timeframe- plus all the devs are tired and need vacations and such. Suddenly, the game is not so portable. It's the same for any platform, really- you go balls to the wall optimizing our game for the platform and you're going to spend a lot of your smooth portability.
Pay no attention to the "specs" of consoles vs. PC, it's basically meaningless. Consoles often run games almost directly, plus they have all sorts of architecture enhancements and little hardware tricks you don't find in PC's. A PC needs to have brutally more power to really match the sort of speed and power you can squeeze out of a console.
Let's say you developed on nintendo wii first... well, it's game over already, you just developed a last-gen, almost Xbox-looking game and tied it to the wiimote. Good luck porting that. That's part of why American studios don't throw big games at it, because it's too limited in power and the publishers just don't want to risk it. There are too many "hardcore" games, which need to push the envelope. The Wii is basically doomed to casual games and childrens' games because of this, because the marketing figures will always point it in that direction--and that's what really runs the game industry.
Technically speaking, you can probably see why people like the Unreal Engine or Source Engine, given the fact that all the porting work is done for you... well you still have to deal with the insane, i mean ABSOLUTELY insane requirements each console has for release... everything from trademarks to menu formats to the way control is expressed in the interface. The amount of attention to detail necessary blows away months of work. Consoles are not a free-for-all, you have to use the hardware in a very specified way.
In short.. yeah, it's rough. More difficult than most people will ever really know.
I confess I'm a game porter, I'm deep into the bowels of finishing off a port of the original Call of Duty to Xbox 360 and PS3 at the moment. Most of the time the ports are outsourced to companies like ours rather than developed in-house by the original developers. We usually have a short development schedule and are pretty much stuck with the code as is, as excellent or crappy as it might be, and we do our best to make what we can from it. I actually find it very intellectually challenging and fun. The schedules are short, and there's always a new project to look forward to while being stuck in the muck of the current project. :) I get to look at a lot of different source code from a lot of different games and learn something new each time usually. Each project is different, sometimes it's easy (if the code is designed well or uses middleware that's available on the platform we're porting to) or a complete nightmare (very platform specific or the middleware it's using isn't available for the platform). At this point I've ported to or from just about every platform out there. Xbox -> PC, PC -> PS3,Xbox, DS -> iPhone, PC -> Mac, etc.
The big flaw in all this is an assumption that any video game publisher wants consoles to be killed.
Some people develop video games but do not do so as a full-time day job. They want consoles to be killed because console makers (especially Nintendo) have an overt bias against teams who work from home. This means games developed in part-time have to be self-published for PC. And even among major labels, there have been a couple stories on Slashdot over the past couple days about publishers whining about console makers' fee structures. See, for example, this story and this story.